18.june.2004 Second assignment : On-line investigation The Reaction Time experiment A comparison between native Persian speakers and native Arabic or English speakers Course EDSL 617 Teaching and learning second language vocabulary Prof.: Thomas Cobb Prepared by : Pirmorad-kandi Code permanent : PIRS08587005 McGill University Introduction In the experiment of the investigation which was done by Randall and Meara over Arabic native speakers who were learning English as second language (1988), the difficulties in processing Roman script and the reading strategies employed for recognising a target letter in a string of Roman letters of Arabic students was compared to the same as native English speakers. The main purpose of this study is to find the way that native Arabic speakers search string of letters and, or any degree of similarity between the search strategies of speakers using the Roman alphabet and those using Arabic script. In their experiment task students were asked to identify a target letter in a string of five characters of Roman letters. For the responses to the target letter for each of the five positions ,they measured their reaction time too. Randall and Meara observed that the target letters were recognized fastest (by Arabic speakers) when they lie in the middle of the array, and more slowly when they lie at the end of the array for both of the Arabic or roman letters. It means that their reaction time to Roman letters was in the same way as to Arabic letters. The form of graphed shape for five positions of these ESL learners was U-shaped. In this study only the random letter strings were used. It might be take different results by using real words already learnt because it takes more attention to read a random word letter by letter than a real word. The results of their experiment over English native speakers showed that the fastest responses belonged to positions in which target letters were on the left of the string. There were slightly faster responses to target letters when they located either in the middle or in the rightmost end of the string. Thus, the graphed shape form was described as a M-shaped By comparing the results of the Arabic native speakers strategies employed for recognising target letters of Roman letters in the string to the same as Arabic letters, Randall and Meara found the main factors which can be affect the using of these strategies in identifying targets in both of these alphabets (Roman and Arabic). The reasons to explain the using of the same strategies due to Arabic systematically differences from English: firstly, from right to left direction of script in Arabic writing, being a consonantal language (the vowels are not been written; according to the context, the correct vowels have to been supplied) and consisting of three-character root (the majority of Arabic words are less than six characters long with 3 letters as the root which locate almost in the middle of each word) In the next experiment, Randall and Meara tried to find the way in which native Arabic speakers with a more sophisticated academic background read Arabic too. They expected to find different results from the results of the first experiment. The results of second study showed essentially the same results as those in the first one. Thus, they concluded that academic sophistication was not a major factor in the effects observed. I tried to investigate this experiment over native Persian language speakers Persian language has approximately the same systematical writing as Arabic language except for three-character root which does not exist in it and four more letters than Arabic. Thus, there are long words with more than five or six characters too; like as English language (vowels are supplied according to the understanding of the context). I was interested in experimenting the same test over Iranian participants. Methodology Participants Two participants were selected for this experiment Behnaz and Narin. Their native language is Persian and, both are from Iran. Behnaz is college student at the first of 2nd year grade and, Narin has studied mainly in French literature in Iran ( learnt in French language) and nursing from a Canadian college (learnt in English language). Narin has had more contact with English than Behnaz. She is 42 years old and Behnaz is 21. Materials For this experiment, the participants were shown a target letter followed by a set of five letters that consisted of 150 English real words. The target letter in each case was “m” which existed in 100 of words 20 items for each of the 5 positions (oooom, ooomo, oomoo, omooo, moooo) . The other 50 words served as distracters that did not contain an ‘m’. The words for the present study were found through the 4 kind of first assignment texts (fiction, conversation, newspaper and academic) and Merriam Webster online dictionary Procedure This test was made by using the Reaction Timer, available in the lexical Tutor Web (http://www.lextutor.ca) for measuring the reaction time calculating standard deviations for each of the five positions and for both participants The experiment was done at researcher’s home. The participants were asked to press key “2” for start , key “1” for “yes” and key “3” for “no“ as quickly and as correctly as possible whether the target letter appeared or not in the strings. The words were appeared randomly. The participants were not given any other instruction about the objective of the experiment. They were asked to do the practice test of 15 items in which they were asked to identify the letter “x“ provided in the lexical tutor web before doing the real test to being ensured that the participants would be comfortable with test. Results The results for reaction time and standard deviations for different positions of the letter “m” of the experiment done with Narin were represented in Table 1 and , graph 1 that is showing her search patter which is N-shaped (maybe). The fastest RT(reaction time) of Narin is noticed in the 4th position (0.7) with 0.111 standard deviation. Narin’s errors rate is low (1.6%). The slowest RT of this participant is in the 3rd position in which the target letter “m” lies in the middle of the strings (0.844) with a standard deviation of 0.144. It shows approximately the same results for the second (0.781) and last position (0.783). The first position results is neither so low as the 3rd nor as high as the forth one (0.762) with 0.115 standard deviation. Table 1. Reaction time. Standard deviations and average scores of five positions of the letter “m” for Narin. “m” letter position 1 moooo 2 omooo 3 oomoo 4 ooomo 5 Oooom Average 0.762 0.781 0.844 0.7 0.783 0.774 0.115 0.17 0.146 0.111 0.206 0.153 Reaction Time St-dev Graph 1 - Reaction time for Narin. 1 0.8 0.6 Series2 Series1 0.4 0.2 0 moooo omooo oomoo ooomo oooom Table 2 and , graph2 show the results for reaction time and standard derivations for different positions of the letter “m” of the experiment done with Behnaz that is noticed as V-Shaped (maybe). The fastest RT(reaction time) of Behnaz is measured in the 2 nd position (0.535) with 0.095 standard deviation. Behnaz’s errors rate is even lower than Narin (1.35%). The slowest RT of this participant is in the 4th position in which the target letter “m” lies in the rightmost end of the strings (0.601) with standard deviation of 0.124. The results of the first (0.592) and 3rd (0.595) position do not show big differences. We observe an increasing RT is noticed from the first to the second position (from 0.595 to 0.535) , in contrast, a decreasing RT from second position to third and then to forth( from 0.535 to 0.595 and 0.601). Table 1. Reaction time. Standard deviations and average scores of five positions of the letter “m” for Behnaz “m” letter position 1 moooo 2 omooo 3 oomoo 4 ooomo 5 oooom Average Reaction Time 0.592 0.535 0.595 0.601 0.588 0.583 St-dev 0.095 0.107 0.128 0.124 0.099 0.113 Graph 1 – RT (Reaction time) for Behnaz 0.62 0.6 0.58 Series1 Series2 0.56 0.54 0.52 0.5 moooo omooo oomoo ooomo oooom Discussion By looking at these results, we can conclude that the two native Persian speakers performed differently in the present experiment. There are number of reasons why their pattern of responding do not give the same shaped curve. Firs, it could be the main factor of age. As it is mentioned in methodology section, Behnaz is much younger than Narin. Secondly, the background of studying in Roman letters, their experience of contact in Roman letters is not same. Narin has studied in Roman letters for 15 years but Behnaz has been studying in Roman letters since last year. It means that Behnaz has a stronger Persian background than Narin. By comparing the RT of these two participants to the RT results of a native English speakers, who had the fastest responses to targets on the left of the string, it is a little similar to Narin’s RT in the first position which was the second fast RT in her test whereas, the fastest RT of Behnaz was in the second position. The lowest responses to target letters of the native English speakers occur when the target letter lies either in the second or in the forth position. Narin’s slowest RT was noticed in the 3rd position of target letter in string that is a semi-similar to the native speakers RT. Now, it is more interesting the comparison between the results of these two native Persian speakers to the RT results of native Arabic speakers because, Persian language script has exactly the same systematical characteristic as Arabic except for three-character, root of each word. Its writing direction is from right to left and a consonantal language. According to Randall and Meara investigation, native Arabic speakers were represented the fastest recognition of target letters if they lie in the middle of array which is the position in which Narin has the lowest RT and, Behnaz had a semi similarity to Native Arabic results in the fastest RT; she identified fastest in second position but, she was not as fast as the Arabic speakers. Arabic speakers were fast when the targets locate in the middle of the string nevertheless; both of the native Persian speakers were slow when the target was in the middle of the string. Conclusion In this experiment, it would be expected to produce a U-shaped search function similar to the function produced by native Arabic speakers. It has been shown that in spite of the similar principles between Arabic and Persian languages, search function is not produced in the same way by two groups ( Arabic and Persian). As a result, the reasons mentioned in the article named as title : “HOW ARABS READ ROMAN LETTERS” , Randall and Meara(1988) why native speakers of Arabic to develop search functions might are not enough for discussing about reading strategies used by Arabs. If we suppose that it could be similar in the second case (Behnaz V_Shaped) , then, what about the third reason of this article which mentions- one of the main characteristics of Arabic three-character root of Arabic language ? This character does not include in Persian language. Limits : This experiment has been done only over two Iranians. It might be necessary more investigation in order to gain the best results not only over some more participants but, on the Persian and Arabic languages similar or different characters as well. References Cobb, T. (2003). The reaction time instrument builder v.1. Retrieved Monday, June 15, 2004, from http://www.er.uqam.ca/nobel/r21270/react_time/ Randall, M. & Meara, P. (1988). How Arabs read romance letters. Reading In A Foreign Language, 4,